HP P6000 Enterprise Virtual Array release notes (XCS 10001000) (5697-1819, March 2012)

• Fixed an issue where back-end transport errors caused the controller software to fail a disk drive
unnecessarily.
• Fixed an issue where the array would not log any events to indicate the back-end disk enclosures
were improperly cabled.
• Fixed an issue where the I/O module firmware upgrade gets blocked when a disk enclosure
disappears and then remains blocked when the enclosure is accessible again.
• Fixed a problem that, in rare circumstances, occurred when I/O to a disk volume failed and a
bad block replacement was not attempted. As a result, the controller could experience a stalled
state and execute a CSM Reset termination to clear the stalled state.
• To eliminate unnecessary service actions, reporting of events 0D DE 95 06: A drive
enclosure I/O module is not communicating with the other I/O module and
09 CD C3 05: SCMI Fibre Channel port FAILED have been discontinued.
• Fixed a rare controller restart that occurs when reading the link status from disk drives while
workloads are in progress.
• Addressed several situations where drives could erroneously be reported as Single Port on Fibre
(SPOF), indicating it is only connected to one of the two device side I/O loops, and has lost
redundancy.
• Fixed an issue where some drive check conditions were not being handled as transport errors,
resulting in drives being failed incorrectly.
• Fixed an issue where background processing can be blocked and cause a CSM Reset termination
if a drive firmware SMART (Self Monitoring Analysis and Reporting) error is encountered while
waiting for a drive quiesce to complete or when accessing disk media.
• Fixed an issue where the disk enclosure shutdown process would return a failure status even if
the shutdown operation succeeded.
• Fixed an issue where some fields were missing or not completely populated in log events related
to disk enclosure link module status or firmware upgrades.
• Fixed an issue where poorly seated back-end cables could result in too many dropped frame
(code 0x061d4709) and metadata error bit (e-bit) set (0x070e0015) events, flooding the logs.
• Corrected an issue where the incorrect WWID entered from the operator control panel of an
EVA6400/8400 was accepted and stored with the secondary controller.
• Fixed an issue where internal parity generation and checking resources were not configured
correctly.
• Added event logging to output the current controller hardware revision and control and cache
memory configuration during boot.
• Fixed event strings related to System Inoperative events.
• Fixed an issue where the array logs a spontaneous power PC reset termination event instead of
the power loss event when power fails during a restart.
• Fixed an issue where a controller host port could stop responding after a restart due to the improper
processing of brief, recovered link failures.
• Fixed a controller restart that resulted from unexpected host mode sense requests.
• Fixed a rare condition involving duplicate entries in a host port transaction buffer time and
organization queue. The problem appears to occur in all types of controllers and is not port
specific. A rare condition results, which may cause the item being aborted to be completed twice.
• Fixed unnecessary failover/failback Unit Attention messages that are sent to hosts during controller
reboot. The messages affect hosts configured with host modes HP-UX, Sun Solaris, Windows
2003, and Windows 2008.
Fixes (XCS 10000000) 7